Simons Collaboration on the Global Brain: Mapping Beyond Space | Simons Foundation Simons Collaboration on the Global Brain: Mapping Beyond Space | Simons Foundation

Annual Report

2017 Edition

Simons Collaboration On The Global Brain: Mapping Beyond Space
As a rat runs along a straight track, different neurons in the rat’s hippocampal region fire. (As the rat enters the green portion of the track, for instance, the green neurons fire.) Researchers later conducted a similar experiment in which rats manipulated a joystick to change the frequency of a tone. The work showed that neurons respond to changes in sound frequency in a manner similar to their response to changes in location.
Illustration adapted from J.W. Rueckemann and E.A. Buffalo/Nature 2017

Whether you’re lost in a new city or driving the most familiar roads, both your hippocampus and your entorhinal cortex are hard at work. Together, these two brain regions create a powerful human navigation system, with diverse cells coordinating to perform different navigational functions, such as tracking the location, speed and direction of your movement. Discovering this navigational system was a huge feat in neuroscience, earning John O’Keefe, May-Britt Moser and Edvard Moser a Nobel Prize in 2014.

But a growing body of research suggests these brain regions may play an even more expansive role in how the brain organizes experience. New findings from David Tank of Princeton University, who is director of the Simons Collaboration on the Global Brain (SCGB), and others show that the cells of the hippocampus and entorhinal cortex can encode much more than physical space: They can also track sound and other factors. And Lisa Giocomo, a neuroscientist at Stanford University and an SCGB investigator, and her collaborators have shown that these cells seem to be capable of rapidly changing their coding properties to respond to a novel environment or task. (Learn more about SCGB here.)

Neuroscientists have traditionally defined cells in the hippocampus and entorhinal cortex according to the specific navigational ability they seem to support. Neurons known as grid cells track general location in space. Place cells encode specific locations, firing whenever you pass your house, for example. Still others act as speedometers, encoding how fast you are moving. “People envisioned the system as a GPS: The brain knows the speed and direction you’re traveling from specific cells and can then determine distance,” Giocomo says.

That approach has drawbacks, however. It requires researchers to hypothesize a priori as to what they believe a cell is doing — such as encoding speed or location — and look for cells that respond to these variables. To find place cells, for example, scientists look for cells that fire most when the animal is in a particular location but grow quiet as the animal moves away. The problem is that most cells in these regions don’t behave in such a predictable way and therefore cannot be labeled with a function.

Giocomo and her collaborators developed an assumption-free method to define cells’ roles. Rather than looking for cells that code for predefined spatial properties, the researchers have developed statistical models that determine which stimuli best predict whether a neuron will fire. “We’re trying to break free of a preconceived notion of how the brain is operating,” Giocomo says. “With our approach, you can take a blind perspective.”

With this system, Giocomo’s team has been able to identify what 75 to 90 percent of cells in the region respond to, compared with less than 50 percent using traditional methods. “We are now capturing a huge percentage of what the entorhinal cortex is doing during behavior,” Giocomo says.

The analysis, published in Neuron in April 2017, revealed that many cells in the entorhinal cortex are both complex and flexible. A cell might code for one property when an animal is running slowly and another when it runs fast. “That’s a fundamentally different way of thinking about how the brain might compute location in space,” Giocomo says. “What the field has defined as grid cells is probably on one end of [the] spectrum — it’s just the tip of the iceberg in terms of coding features.”

“We’re trying to break free of a preconceived notion of how the brain is operating.”

Dmitriy Aronov of Columbia University and David Tank have also found striking flexibility in both the entorhinal cortex and the hippocampus. The team showed in 2014 that animals learning to navigate a virtual reality world shifted their internal maps when experimenters shifted that virtual world. They reasoned that if animals can readily adapt to a new spatial reality, maybe they can map other types of experiences as well.

To test that theory, Aronov, then a postdoctoral researcher in Tank’s lab, trained rats to traverse an auditory rather than a physical space. The animals used a joystick to move through a sort of soundscape — a defined sequence of frequencies. Moving the joystick to the left, for example, might increase the frequency.

The researchers discovered a set of cells that act very much like place cells. Instead of firing when the animal is in a specific location, these ‘sound cells’ fire when the animal hears a specific tone. The findings were published in Nature in March 2017.

Tank, Aronov and others have also found that neurons in the hippocampus seem to respond not just to space or sound, but to every aspect of the task, firing in a predictable sequence throughout an experiment. “From pressing the lever to traversing the frequency to receiving the reward and starting a new trial — there is a sequence of activation throughout the entire period of behavior,” Tank says.

Taken together, the research suggests that cells in the hippocampus and entorhinal cortex are far more flexible than scientists thought. “The findings point to a more general function of the hippocampus,” says Aronov. Rather than simply encoding where an animal is or how fast it’s moving, the neurons in this brain region map out whatever variables seem to be most important in that context. “If the animal is navigating through space, the sequence [of electrical activity in this group of neurons] tends to correspond to space. If it’s traveling through sound, the sequence will correspond to successive sounds.”

Researchers theorize that this flexibility, also known as remapping, helps animals encode experiences much more broadly. Remapping helps the brain “distinguish between experiences with strongly overlapping elements, such as parking your car in the same parking lot but in different parking spots each day,” Giocomo says.

Giocomo and Tank are now working together, along with Surya Ganguli, Loren Frank, Elizabeth Buffalo, Uri Eden and Ila Fiete, on a new SCGB project that will delve more deeply into the remapping process. Their project will expand on previous research, which focused on how neurons remap in response to visual changes in the environment — if the wall color in a room changes or a wall is knocked down, for example. Giocomo and Tank’s work shows that remapping can happen quickly, such as when an animal changes its behavior, and across different modalities, such as sound. “We want to use that as a springboard to understand how neurons remap, what time course this remapping follows and how this information then gets communicated to the rest of the brain to form a memory or drive behavior,” Giocomo says.

Giocomo, Tank and their collaborators plan to record electrical activity from the same neurons as animals perform two different tasks — foraging for food in an open environment and finding their way through a maze — and analyze how neuronal activity differs depending on the environment and the task. They aim to uncover how remapping helps the brain encode unique experiences and generalize across different experiences.

Advancing Research in Basic Science and Mathematics Subscribe to our newsletters to receive news & updates